S100a6 knockdown promotes the differentiation of dental epithelial cells toward the epidermal lineage instead of the odontogenic lineage

Tooth development is a complex process involving various signaling pathways and genes. Recent findings suggest that ion channels and transporters, including the S100 family of calcium‐binding proteins, may be involved in tooth formation. However, our knowledge in this regard is limited. Therefore, this study aimed to investigate the expression of S100 family members and their functions during tooth formation. Tooth germs were extracted from the embryonic and post‐natal mice and the expression of S100a6 was examined. Additionally, the effects of S100a6 knockdown and calcium treatment on S100a6 expression and the proliferation of SF2 cells were examined. Microarrays and single‐cell RNA‐sequencing indicated that S100a6 was highly expressed in ameloblasts. Immunostaining of mouse tooth germs showed that S100a6 was expressed in ameloblasts but not in the undifferentiated dental epithelium. Additionally, S100a6 was localized to the calcification‐forming side in enamel‐forming ameloblasts. Moreover, siRNA‐mediated S100a6 knockdown in ameloblasts reduced intracellular calcium concentration and the expression of ameloblast marker genes, indicating that S100a6 is associated with ameloblast differentiation. Furthermore, S100a6 knockdown inhibited the ERK/PI3K signaling pathway, suppressed ameloblast proliferation, and promoted the differentiation of the dental epithelium toward epidermal lineage. Conclusively, S100a6 knockdown in the dental epithelium suppresses cell proliferation via calcium and intracellular signaling and promotes differentiation of the dental epithelium toward the epidermal lineage.


| INTRODUCTION
Tooth development involves the entry of the oral epithelium, epithelial-mesenchymal interactions, differentiation of the dental epithelium into ameloblasts, and the formation of the enamel.In mouse, similar to that of hair and nails, formation of incisors is a continuous process.Therefore, given that immature cells are present on the root side and highly differentiated cells are present on the crown side of the tooth, a study of mouse incisors may improve our understanding of the differentiation process of tooth. 1 Moreover, analysis of the labial side of mouse incisors is an invaluable tool for observing the differentiation process of the dental epithelium. 1,2Recent advances in single-cell RNA-sequencing have facilitated mRNA analysis at the single-cell level, enabling the identification of molecules expressed in each cell.Ameloblasts are not monophasic and express different molecules depending on the stage of differentiation; expressing dentin sialophosphoprotein (Dspp) in the early stage of differentiation and strongly expressing ameloblastin (Ambn) in differentiated ameloblasts. 3][6][7] Calcification is performed using the enamel matrix as a scaffold, 4,5 which is later degraded by proteases such as matrix metalloproteinase-20 and kallikrein-related peptidase 4 (Klk4) to enhance calcification.Although extensive studies have been performed on the differentiation of the dental epithelium into mature ameloblasts, 8,9 the calcification process remains poorly understood.
Ion channels play important roles in tooth calcification.Ion channels and transporters are involved in enamel, dentin, and root formation, and in tooth eruption. 102][13][14][15][16][17] Overall, these findings indicate that changes in calcium ions are important for enamel calcification.The S100 family of calciumbinding proteins regulates intracellular calcium concentrations, and is expressed during tooth formation as determined in our previous preliminary experiments.We previously reported that Sox21-deficient mice lack tooth enamel. 18The dental epithelial cells differentiate into hair cells without differentiating into ameloblasts.Therefore, expression of not only Ambn and Amelx but also Klk4 and S100 calcium-binding protein a6 (S100a6) was suppressed.This suggests that S100a6 may be involved in tooth development.S100a6, the expression of which is increased in tumors, promotes epithelialmesenchymal transition (EMT) in pancreatic and breast cancer cells. 19,20Furthermore, proteome analysis suggested that S100a6 is important for EMT. 21However, the calcium-binding protein S100A6 accelerates the growth of human osteosarcoma mesenchymal cells by promoting cell proliferation and inhibiting osteogenic differentiation. 22These results suggest that S100a6 overexpression in the epithelium may cause EMT and alter cell lineage by changing the direction of cell differentiation.However, the role of S100a6 in tooth development has not been elucidated yet.Therefore, this study aimed to investigate the expression of S100 family members and their functions during tooth formation.

| Animals
C57BL/6J mice were used as wild-type (WT) mice.The Tg(KRT14-RFP)#Efu (Krt14-RFP) mouse line was obtained from Dr. Matthew P. Hoffman and maintained as a homozygous line of FVB/N. 23Sox21 knockout (Sox21 KO) mice were provided as gifts from Prof. Hideyuki Okano.The Sox21 coding sequence in C57BL/6 mouse was replaced with EGFP cDNA. 24All animal experiments were approved by the Animal Ethics Committee of the Tohoku University (approval number: 2020DnA-016-05), and performed in accordance with laws on the conservation and sustainable use of biological diversity.Sox21 knockout (KO) mice were fed a soft diet containing 3% calcium in CMF purchased from Oriental Yeast Co. Ltd. (Tokyo, Japan).All genetic recombination experiments were approved by the Tohoku University Center for Gene Research (approval number: 2020DnLMO-003-02).

| Microarray analysis
Tooth germ was extracted from C57BL/6J mice on days 11, 13, and 14 of the embryonic stage.Skin, gingiva, and tooth germs were isolated from post-natal mice on day 1 (P1).Each tissue sample was homogenized using stainless steel beads and TOMY Micro Smash (MS-100).RNA was extracted from the crushed tissue using QIAGEN RNeasy Mini Kit (74104).The samples with RNA integrity number more than nine were used for analysis.The RNA samples labeled using Cyanine 3 (Cy3) dye were hybridized onto SurePrint G3 Mouse GE 8× 60 K microarrays (Agilent Technologies, Santa Clara, CA, USA).SF2 cells and SF2 cells with siRNA-mediated S100a6 suppression were isolated.The labeled RNA samples were hybridized onto SurePrint G3 Rat GE 8× 60 K microarrays (Agilent Technologies).Microarray data were analyzed using GeneSpring software, V13.1 (Agilent Technologies).All gene expression data are presented as volcanic plots.The total number of detected entities was filtered using the signal intensity value to remove extremely low-signal entities.Normalization was performed using a percentile shift in the analysis of signaling pathways.Genes substantially altered by si100a6 knockdown were annotated in the gene ontology (GO) database.
7][28] Thereafter, 12 212 cells from the P1 molar and 6260 cells from the P7 incisor datasets were used for subsequent analyses.Secondary analysis of scRNA-seq was performed using Loupe Browser 6.5.0 (10× genomics), 29 whereas clustering and differential expression analyses were performed as previously described. 28Clustering was performed using graph-based algorithms.The p-values were adjusted using the Benjamini-Hochberg correction for multiple tests.Genes with log2-fold change >1.0, and p < .1 were classified as differentially expressed genes (DEGs).

| RT-qPCR
Total RNA was extracted using Qiagen RNase Kit according to the manufacturer's instructions, followed by reverse-transcription of the RNA into cDNA using Superscript Vilo Master Mix (11756050, Thermo Fisher Scientific).Quantitative PCR was performed using the Applied Biosystems Step One Real-Time PCR system (Thermo Fisher Scientific) with Applied Biosystems SYBR Select Master Mix (4472920, Thermo Fisher Scientific) and specific primer pairs.The primers used are listed in Table S1.The results are representative of at least three independent experiments with triplicates per experiment.

| Live imaging of intracellular calcium and endoplasmic reticulum (ER) fluorescent staining
Briefly, SF2 cells were cultured on chamber slides (154526PK, Thermo Fisher Scientific) with or without NT-4 and Ca 2+ .After two PBS washes, cells were stained with Cell Navigator Live Cell Endoplasmic Reticulum Staining Kit (22636, AAT Bioquest, Inc., Sunnyvale, CA, USA) in the dark at 37°C for 30 min, washed three times with PBS, and further incubated in PBS containing 4 μM Fluo-8 (21080, AAT Bioquest, Inc.) and 2 μM Hoechst 33285 in the dark at 37°C for 60 min.After two washes in PBS, the stained areas were identified using confocal microscopy (Fluoview Fv10i LIV; Olympus).Calcium levels in cells, cultured in 96-well plates, were also determined using Fluo-8 staining by the same method (n = 3) and measured using a TriStar 2 LB 942 plate reader (Berthold Technologies GmbH & Co.KG, Bad Wildbad, Germany).Fluo-8 fluorescence intensity was normalized using 1 mM Ca 2+ -treated cells without NT-4 and si100a6 cells.

| Cell proliferation assay
Briefly, SF2 cells were cultured on a cover glass in Falcon 6-well plates (Thermo Fisher Scientific).Cell proliferation was evaluated using the Click-iT Plus EdU Cell Proliferation Kit for Imaging (C10639, Thermo Fisher Scientific).Cells were incubated in a medium containing 10 μM EdU for 30 min, washed twice with PBS, incubated with paraformaldehyde for 15 min, and washed twice with PBS.Fixed cells were incubated in Click-iT Plus reaction solution in the dark for 30 min and washed twice with PBS.For nuclear staining, cells were incubated with 2 μM Hoechst 33285 in the dark for 30 min, washed with PBS, and viewed using a confocal microscope (Fluoview Fv10i LIV; Olympus).

| Statistical analyses
The data are presented with mean values shown as bars.Error bars represent the standard deviation.A two-tailed Student's t-test was used for the statistical analyses of two independent variables.Statistical significance was set at p < .05.

F I G U R E 1
Microarray analysis was performed to determine the expression of S100 family members in embryonic mouse tooth germ.(A) Changes in the expression of S100 family members in embryonic mouse tooth germ on days 11 (E11), 13 (E13), and 14 (E14).Heatmap showing the expression levels of genes, with red color representing upregulated genes and white color indicating downregulated genes.Fold change was calculated from expression on E13/E11 and E14/E11.(B) Single-cell RNA-sequencing (scRNA-seq) was performed to determine the expression of S100 family members in post-natal mouse molars on day 1 (P1).(C) Expression of S100 family members in post-natal mouse incisor on day 7 (P7).Mouse molar and incisor tooth germs were extracted.Isolated tooth germ cells were analyzed using scRNA-seq.Clustering was performed based on the gene expression pattern of each cell.(D) The expression of S100 family members in various tooth germ cells.The size of the circle is the percentage of expressed cells.Color intensity for average expression was calculated as the average of the normalized expression values of all cells within each class, and then scaled to have a mean of zero and a standard deviation of one.

| Subgroup A of the S100 family is uniquely expressed in the tooth germ
Microarray analysis was performed to investigate the expression of S100 family members in embryonic mouse tooth germs (Figure 1A).S100a10 and S100a11 were strongly expressed in embryonic tooth germs on days 11, 13, and 14.In contrast, S100a3, a4, a5, a7a, b, g, and pbp showed low expression levels at all stages.Additionally, S100a1, a6, a8, a9, and a14 expression levels increased during tooth differentiation.Particularly, S100a6, a8, and a14 expression increased by 6.8-, 8.0-, and 4.7-fold, respectively, suggesting that these genes are crucial in the later stages of tooth differentiation.
scRNA-seq was performed on mouse incisor and molar tooth germ.S100a6, a10, and a11 were highly expressed in ameloblasts in post-natal mouse molar on day 1 (P1; Figure 1B).Non-ameloblasts, inner enamel epithelium (IEE), and ameloblast clusters are dental epithelial lineages.Given that S100a10 and S100a11 were also strongly expressed in mesenchymal cell clusters, it could be concluded that they were also expressed in the entire P1 molar.S100a14 and S100a16 were weakly expressed in non-ameloblasts, IEE, and ameloblasts, but not in mesenchymal cells.Overall, S100a6, a14, and a16 showed higher expression in epithelial cells than mesenchymal cells in P1 molars.Moreover, S100a1, a6, a8, a9, a10, a11, a13, and a16 were highly expressed in ameloblasts in the incisor on day 7 (P7; Figure 1C).Notably, S100a8, a9, a10, a11, a13, and a16 were also expressed in mesenchymal cell clusters, but not expressed specifically in the P7 incisor.S100a14 was weakly expressed in ameloblasts, but not in mesenchymal cells.Collectively, S100a1, a6, and a14 showed higher expression in epithelial cells, but not in mesenchymal cells in P7 incisors.Single-cell analysis of P1 molars and P7 incisors suggested that S100a6 and S100a14 F I G U R E 2 Immunostaining for S100a6 expression in mouse molars.Mouse molars were immunostained with an anti-S100a6 antibody (green), and nuclei were stained with Hoechst 3328 (blue).S100a6 was detected at an excitation wavelength of 473 nm and an emission wavelength of 520 nm using a confocal microscope.Hoechst was detected at an excitation wavelength of 405 nm and an emission wavelength of 455 nm using a confocal microscope.(A) The boundary between the epithelium and the mesenchyme is indicated using a dotted line.Low magnification images of embryonic mouse molars on days 13 (E13) and E15, and post-natal mouse molars on days 1 (P1), P7, and P14.(B) High magnification images of the area outlined by the white square in the low magnification images on P1, P7, and P14.Scale bar = 200 μm.AM, apical membrane side; BM, basal membrane side.

F I G U R E 3
Immunostaining for S100a6 in incisors from wild-type (WT) and Sox21-deficient mice.Incisors of WT and Sox21KO mice were immunostained with anti-S100a6 antibody (green), and nuclei were stained with Hoechst 3328 (blue).S100a6 was detected at an excitation wavelength of 473 nm and an emission wavelength of 520 nm using a confocal microscope.Hoechst was detected at an excitation wavelength of 405 nm and an emission wavelength of 455 nm using a confocal microscope.were specifically expressed in ameloblasts.Additionally, S100a10 and S100a11 were strongly expressed, but were not specific to time or cell type (Figure 1D).Microarray and single-cell analyses showed that among the genes examined, S100a6 had the highest fold change and expression in ameloblasts, respectively; therefore, it was selected for further analyses.

| S100a6 is expressed at the apical membrane side in secretory stage ameloblast
Immunostaining was performed to investigate S100a6 expression in mouse molar tooth germ (Figure 2).S100a6 was not expressed in embryonic molar tooth germs (Figure 2A,B) and undifferentiated dental epithelium.In contrast, S100a6 expression was detected in ameloblasts and stellate reticulum in post-natal molar tooth germ on day 1 (P1).However, S100a6 expression showed a decreasing trend in the stellate reticulum in post-natal molar tooth germ on days 7 and 14, but was strongly expressed in ameloblasts.Additionally, S100a6 expression was localized to the apical membrane side of the cell to initiate the formation of the enamel (Figure 2B).The enamel is located only on the labial side of the incisors.Given that the enamel continues to erupt, undifferentiated dental epithelial cells were present on the root side, whereas differentiated mature ameloblasts were present on the crown side.Therefore, we examined the expression of S100a6 in ameloblasts in the incisors of WT mouse (Figure 3A-D).S100a6 expression was observed in ameloblasts that differentiated from the dental epithelium, with the expression level increasing with the progression of differentiation (Figure 3C,D).
Sox21-deficient mice lack enamel, and are used as a model mouse for amelogenesis imperfecta. 18S100a6 mRNA expression was also decreased in Sox21-deficient mice and in SF2 cells in which Sox21 expression was suppressed by siRNA (Figure S1).Therefore, protein expression of S100a6 in Sox21-deficient mice was investigated via immunostaining.Similar to molar ameloblasts, S100a6 was localized to the apical side of the cells.In Sox21KO mice, ameloblasts showed loss of polarity and became multilayered, which inhibited enamel matrix degradation (Figure 3E-H, arrowhead).S100a6 suppression was also observed in ameloblasts during the secretory phase; and it was expressed in the nuclei of cells (Figure 3G,H, arrow).Overall, these results indicate that S100a6 expression is suppressed in enamel hypoplasia.
An in vitro experiment was performed using a rat dental epithelial cell line (SF2 cells) to investigate the subcellular localization of S100a6.Cell adhesion was unaffected after culturing for 24 h in Ca 2+ -free medium (data not shown).Additionally, S100a6 was localized to the nucleus in a medium containing 1 mM calcium (Figure 4A).In contrast, we observed an increase in S100a6 translocation into the cytoplasm and an increase in its expression therein with increasing concentration of calcium.Moreover, S100a6 was localized to the intranuclear region in unstimulated SF2 cells, but translocated into the interior of E-cadherin, a plasma membrane marker genes, and was, therefore, present in the cytoplasm in NT-4-treated differentiated cells (Figure 4B).Collectively, these results suggest that S100a6 is involved in enamel calcification, as it is locally expressed on the calcifying side of ameloblasts.

| S100a6 suppression in ameloblasts inhibits enamel matrix expression and upregulates epithelial marker genes
To elucidate the role of S100a6 in enamel calcification, we performed siRNA-mediated S100a6 knockdown in SF2 cells.S100a6 knockdown inhibited the expression of enamel matrices Ambn and Amelx but increased the expression of Klk4 (Figure 5A).Additionally, NT-4-induced F I G U R E 4 Immunostaining for S100a6 in SF2 cells.SF2 cells were immunostained with an anti-S100a6 antibody (green), and the nuclei were stained with Hoechst 3328.S100a6 was detected at an excitation wavelength of 473 nm and an emission wavelength of 520 nm using a confocal microscope.E-Cadherin (E-Cad) was detected at an excitation wavelength of 559 nm and an emission wavelength of 618 nm using a confocal microscope.Hoechst was detected at an excitation wavelength of 405 nm and emission wavelength of 455 nm using a confocal microscope. of SF2 cells increased the expression of Ambn, Amelx and Klk4.Moreover, S100a6 inhibition suppressed Ambn and Amelx upregulation in differentiated SF2 cells, but did not affect Klk4 expression.Furthermore, we examined the effect of calcium treatment (1/2/3 mM) on S100a6 expression and enamel calcification.Although treatment with 1 or 2 mM Ca 2+ did not significantly affect the expression of S100a6 and ameloblast marker genes, the parameters were significantly upregulated following treatment with 3 mM Ca 2+ (Figure 5B).siRNA-mediated S100a6 knockdown markedly inhibited Ambn and Amelx expression at all Ca 2+ concentrations, but did not affect Klk4 expression.Moreover, S100a6 knockdown increased the expression of the epithelial marker genes Nectin 2, Nactin 4, Keratin 8, Krueppel-Like Factor 4, and E-cadherin (Figure 5C), and decreased the expression of the mesenchymal marker genes Vimentin and Zeb1 (Figure 5D).

| S100a6 knockdown in ameloblasts inhibited Ca 2+ influx into the cytoplasm
Immunofluorescence assay showed that Ca 2+ was mainly localized to the nuclear membrane and ER in SF2 cells that were cultured in medium containing 1 mM Ca 2+ (Figure 6A).However, treatment with 2-3 mM Ca 2+ increased Ca 2+ translocation from the nucleus to the cytoplasm, with Ca 2+ mainly localized in the cytoplasm (Figure 6A).Similarly, Ca 2+ was mainly detected in the cytoplasm in NT-4-stimulated SF2 cells (Figure 6B).However, a decrease was noted in intracellular Ca 2+ levels in S100a6-knockdown cells.Therefore, the amount of intracellular Ca 2+ was measured (Figure 6C).The results showed that the amount of Ca 2+ increased by 160% in the cells with NT-4.In contrast, inhibition of S100a6 resulted in a decrease in intracellular Ca 2+ .Collectively, these results suggest that Ca 2+ is mainly stored in the ER of dental epithelial cells, but migrates to the cytoplasm during differentiation into enamel blast cells.Additionally, S100a6 is important for the maintenance of intracellular calcium levels.F I G U R E 5 Expression of ameloblast, epithelial, and mesenchymal marker genes in dental epithelium.The expression of ameloblast marker genes, such as Ameloblastin (Ambn), Amelogenin (Amelx), and Kallikrein-related peptidase (Klk4), in SF2 cells was evaluated using qPCR.(A) Ameloblasts differentiated from NT-4treated SF2 cells.S100a6 expression was suppressed by siRNA.The cells were cultured in DMEM/F12 medium containing 2 mM Ca 2+ .(B) SF2 cells were cultured in calcium-and serum-free medium (Keratinocyte SFM) containing 1, 2, or 3 mM Ca 2+ .(C, D) Expression of epithelial and mesenchymal marker genes in SF2 cells cultured with NT-4 or S100a6 siRNA.(C) Expression of epithelial marker genes such as Nectin2, Nectin4, keratin8 (Krt8), Krueppel-Like Factor 4 (Klf4), and E-cadherin (E-cad).(D) Expression of mesenchymal marker genes such as N-cadherin (N-cad), Snail1 Vimentin (Vim), Zeb1, and Zeb2.*p < .05compared to control sample (n = 3).

| S100a6 knockdown inhibits NT- induced activation of ERK signaling in ameloblasts
NT-4 directly activates ERK pathway in SF2 cells to promote ameloblast differentiation. 25Therefore, we examined the effects of S100a6 expression on NT-4-induced activation of ERK signaling in SF2 cells.Expectedly, NT-4 stimulation enhanced the phosphorylation of the ERK pathway in SF2 cells 5-15 min (Figure 7A), which was suppressed by S100a6 knockdown.][32] We investigated the effects of S100a6 on ERK and related proteins.S100a6 knockdown suppressed the phosphorylation of ERK, MEK, PI3K, and AKT in SF2 cells within 24 h (Figure 7B).Moreover, phosphorylation of the proteins was inhibited in NT-4-stimulated cells, suggesting that S100a6 regulates several cellular signaling pathways involved in ameloblast differentiation.In contrast, S100a6 knockdown did not affect the phosphorylation of p38 and Smad2.Overall, these results suggest that S100a6 knockdown not only inhibits NT-4-induced activation of the ERK-MEK pathway, but may also inhibit the PI3K-AKT pathway.

| S100a6 in ameloblasts regulates cell cycle and dermal differentiation
To further elucidate the role of S100a6 in ameloblasts, microarray and differential expression analyses were Intracellular calcium was detected using Fluo-8.Fluo-8 was detected at an excitation wavelength of 473 nm and an emission wavelength of 520 nm using a confocal microscope.ER-stain was detected at an excitation wavelength of 559 nm and an emission wavelength of 618 nm using a confocal microscope.Hoechst was detected at an excitation wavelength of 405 nm and an emission wavelength of 455 nm using a confocal microscope.(A) SF2 cells were cultured in media containing different concentrations of Ca 2+ .(B) SF2 cells were cultured with NT-4 or S100a6 siRNA in DMEM/ F12.SF2 cells were double-stained for endoplasmic reticulum and Ca 2+ to identify the intracellular calcium localization.Scale bar = 20 μm.(C) SF2 cells were cultured with DMEM/F12 supplemented with NT-4 or S100a6 siRNA in 96-well plates.Ca 2+ levels were determined after 24 h using Fluo-8 staining, and the amount of Ca 2+ was measured using a plate reader.*p < .05compared to control sample (n = 3).
to examine the effects of S100a6 knockdown on certain genes.Compared with the control group, S100a6 knockdown downregulated 2666 DEGs and upregulated 3038 DEGs (Figure 8A; Table S2).Overall, these results suggest that S100a6 may act as an expression suppressor.Additionally, pathway enrichment analysis indicated that S100a6 may affect cell cycle, inflammation, and metabolism (Table S3).Consistent with the results of the intracellular signaling analysis (Figure 7), pathway analysis suggested that S100a6 may also affect the PI3K, AKT, and MAPK signaling pathways.GO analysis showed decreased expression of nuclear-and chromosome-related, and cell cycleassociated molecules (Figure 8B).In contrast, molecules involved in skin, keratinocyte, and epidermal differentiation were significantly upregulated.Collectively, these results suggest that S100a6 knockdown in ameloblasts may reduce cell proliferation and promote differentiation toward the epidermal lineage.

| S100a6 knockdown in ameloblasts reduces cell proliferation
GO analysis indicated that S100a6 knockdown may inhibit the cell cycle (Figure 8B); therefore, we examined the effect of S100a6 knockdown on the proliferation of SF2 cells.EdU uptake and SF2 cell proliferation were suppressed following NT-4 treatment (Figure 9A).Similarly, S100a6 knockdown significantly suppressed EdU uptake and SF2 cell proliferation after NT-4 treatment for 24 h (Figure 9B).Notably, the number of EdU-positive cells decreased to one-tenth the initial population after 24 h of S100a6 knockdown.NT-4 treatment and S100a6 knockdown significantly decreased cell proliferation within 72 h (Figure 9C).Collectively, these results suggest that S100a6 promotes cell proliferation.

| S100a6 knockdown in dental epithelium promotes epidermis differentiation
Tooth germ, skin, and gingiva samples were collected from post-natal mouse on day 1, and differential expression analysis was performed following S100a6 knockdown (Figure 10).Upregulated genes in the skin and gingiva compared with those in the tooth germ are shown in Figure 10A.Microarray analysis and RT-PCR were used to determine the expression of upregulated genes in SF2 cells following S100a6 knockdown.The results suggest that S100a6 knockdown promotes SF2 cell differentiation into non-odontogenic epithelial lineages.Compared with that in the tooth germ, several genes were upregulated in the gingiva and downregulated in the skin (Figure 10B).Microarray analysis indicated that expression of most of F I G U R E 7 Analysis of intracellular signaling in ameloblasts following siRNA-mediated S100a6 knockdown.SF2 cells were cultured with S100a6 or negative control siRNA for 24 h.(A) Transfected SF2 cells were cultured with NT-4 for 0, 5, 10, 15, 30, or 60 min.Cytoplasmic components were stained with anti-ERK and anti-p-ERK antibodies.(B) Transfected SF2 cells were cultured with NT-4 for 24 h.Cytoplasmic components, excluding the nuclei, were extracted and stained with specific antibodies for western blotting.
the genes remained unaffected.Collectively, results indicate that S100a6 knockdown in the dental epithelium induces changes in gene expression pattern similar to that in the epidermis.

| DISCUSSION
Cells use various regulatory mechanisms to maintain homeostasis, including the ion channel, which regulates ion gradients and action potentials, ion flow, and cell volume.Ca 2+ functions as a secondary messenger in the regulation of various cellular processes. 33Ca 2+ -binding proteins bind to calcium ions to improve their availability when required.Ca 2+ -binding proteins are mostly cytoplasmic, acting as Ca 2+ buffer or serving as transducers of Ca 2+ signaling.Calmodulin, troponin C, and most S100 proteins are considered Ca 2+ signaling proteins, and all of these have a conserved calcium-binding motif termed EFhand.The S100 family, with two EFs, is divided into three subgroups and is specifically expressed in tissues and cells.However, each molecule is expressed in different tissues, and the site of expression can be intracellular, in the plasma membrane, or extracellular in the secretory form.6][47] In the present study, S100a1, a6, a8, a9, a10, a11, a13, a14, and a16 were expressed in the tooth germ (Figure 1); F I G U R E 8 Microarray analysis for S100a6 expression.SF2 cells and S100a6knockdown SF2 cells were compared using microarray analysis.(A) Volcano Plot: genes with p < .05 and expression ratios >2fold (red) or <0.5fold (blue) were extracted.(B) Functional annotation of differentially expressed genes (DEGs) was performed using the gene ontology (GO) database.Upregulated and downregulated functions are listed.
however, their roles in tooth differentiation were unclear.S100a11 is expressed in the cartilage and promotes hypertrophic chondrocyte differentiation. 48Therefore, it could be speculated that S100a11 may regulate tooth calcification.In the present study, S100a14 was weakly expressed in the dental epithelium 1).S100a14 regulates the expression of MMPs in oral squamous cell carcinoma cells, 49 and enamel calcification is enhanced following the degradation of the matrix scaffold by MMPs.Therefore, S100a14 may contribute to enamel hypercalcification via MMPs.S100a6, also called calcyclin, mainly forms homodimers and binds to two Ca 2+ . 50,513][54] Additionally, S100a6 is detected in MC3T3-E1 osteoblasts, and mediates the cation response. 55Moreover, S100a6 overexpression significantly increased the mRNA expression of alkaline phosphatase in MC3T3-E1 cells. 56Unlike other tumors, osteosarcoma inhibits metastasis and migration. 57In the present study, S100a6 was not detected in immature dental epithelium, but was highly expressed in ameloblasts during differentiation (Figures 2-4).Despite S100a6 expression in ameloblasts, the incidence of ameloblastoma is low among all tumors.Therefore, S100a6 inhibits tumorigenesis by inducing mineralization in teeth and bones, and induces tumorigenesis in organs where calcification does not occur.
The intracellular expression of S100a6 was localized to the apical side that forms the enamel, but not in the basement membrane side (Figures 2 and 3).Generally, S100a6 is mainly localized in the cytoplasm; however, increased calcium concentration promotes the translocation of Ca 2+ to the plasma and nuclear membranes. 58,59imilarly, other ion channels associated with calcification, such as Ncx, Nckx, Stim, and Slc, are expressed in ameloblasts.Moreover, Ncx and Nckx, are expressed on the apical side of ameloblasts to regulate calcium. 10ther calcium-binding phosphoproteins are secretory calcium-binding phosphoproteins (SCPPs) that regulate enamel and dentin calcification. 60Pinctada fucata mantle gene 1 (Pfmg1), found in the epithelium, has two EF hands and a calcium-binding domain that promotes progenitor osteoblast differentiation, and enhances matrix mineralization. 61Therefore, proteins with EF hands, such as members of the S100 family, may contribute to tooth calcification by binding to calcium.In the current study, calcium was present in the nuclear membrane and endoplasmic reticulum in unstimulated SF2 cells.However, treatment with 2-3 mM of calcium and NT-4 stimulation increased the translocation of Ca 2+ into the cytoplasm (Figure 6).Inhibition of Tric-A and Tric-B, which induce calcium release from the endoplasmic reticulum, results in bone dysplasia. 62An increase in extracellular calcium concentration induces dental pulp cell differentiation and enhances calcification.Integrin β1 (Itgβ1) is a cell adhesion molecule that transmits the extracellular effects of S100a6. 63For example, ameloblast cystic formation and differentiation was suppressed in Itgβ1 conditional KO mice. 64Collectively, these results suggest that S100a6 in the plasma membrane regulates enamel formation via calcium binding.1][32] In contrast, in this study, siRNA-mediated knockdown in S100a6 in SF2 cells not only inhibited ERK and AKT activation, but also suppressed NT4-induced ERK activation (Figure 7).

F I G U R E 9
Effects of S100a6 suppression on the proliferation of SF2 cells.(A) SF2 cells were cultured with NT-4 or S100a6 siRNA for 24 h and then stained with EdU (red).The nuclei were stained with Hoechst 33285 (blue).EdU was detected at an excitation wavelength of 559 nm and an emission wavelength of 618 nm using a confocal microscope.Hoechst was detected at an excitation wavelength of 405 nm and an emission wavelength of 455 nm using a confocal microscope.Scale bar = 20 μm.(B) The proportion of EdU-positive cells was calculated from the number of all cells in the field that were incubated for 24 h.(C) SF2 cells were cultured with NT-4 or siS100a6 for 24, 48, and 72 h.The relative proliferation rates were calculated based on the number of EdUpositive cells at 24 h.*p < .05compared to control sample (n = 3).

(A) (B) (C)
Increased extracellular calcium in osteoblasts enhances by extruding calcium from the endoplasmic reticulum via pannexin 3 (Panx3), which is an endoplasmic reticulum Ca 2+ channel, via AKT signaling. 65Cx43 is an intercellular channel that permits the exchange of small ions and secondary messengers, and activates ERK in ameloblasts. 13Therefore, these channels are thought induce calcification not only through ion transfer but also through various intracellular signals.
Matrices such as Ambn and Amelx are important for the initiation of calcification, whereas enzymes such as Klk4 are important for hypercalcification.Therefore, an appropriate dose of protease is essential for enamel formation, and deficiency or excess of Klk4 causes enamel F I G U R E 1 0 Effects of S100a6 suppression on skin marker genes in ameloblasts.Tooth germ, skin, and gingiva were excised from post-natal mice on day 1 (P1) for microarray analysis.Upregulated and downregulated genes in mouse tooth are shown in blue and red, respectively.(A) Significantly upregulated genes in the gingiva and skin compared with the tooth germ.(B) Upregulated genes in the gingiva, but not in the skin, compared with that in the tooth germ.qPCR was performed to detect S100a6 expression in SF2 cells transfected with siRNA or negative control siRNA, and the expression was expressed as Log2 fold change (FC).
(A) (B) dysplasia. 8,66In the present Ambn, Amelx, and Klk4 were upregulated during ameloblast differentiation; however, S100a6 knockdown suppressed Ambn and Amelx expression (Figure 5A).The suppression of Ambn and Amelx suggests that the loss of S100a6 inhibited the differentiation of ameloblasts.Notably, Klk4 expression was not suppressed by the depletion of S100a6 in dental epithelial cells.Ambn and Amelx are not expressed in skin, but Klk4 is expressed in the epidermis. 67The reduced expression of Ambn and Amelx, and unchanged expression of klk4, may have differentiated the dental epithelium into dermal-like cells.9][70][71] S100a6 knockdown increased the expression of epithelial marker genes and decreased the expression of mesenchymal marker genes in ameloblasts (Figure 5C).Additionally, GO analysis indicated increased expression of molecules involved in skin, keratinocyte, and epidermal differentiation (Figure 8B).The results of PCR indicated that S100a6 has similar mRNA expression pattern in ameloblasts and skin tissue (Figure 10); therefore, it was speculated that S100a6 knockdown may not only suppress ameloblast differentiation, but promote their differentiation into the epidermal lineage.S100a6 suppression also upregulated S100a4, a7a, a8, and a9 expression (Table S1).S100a4 and S100a7a were not expressed in the tooth germ, whereas S100a8 and a9 were weakly expressed in the mouse incisors (Figure 1).S100a4 is expressed in inflammatory and tumor cells; it induces an increase in matrix-degrading enzymes, and is involved in rheumatoid arthritis. 72S100a7a is expressed in keratinocytes, and is enhanced during inflammatory response. 73S100a8 and S100a9 form heterodimers, and are upregulated in tumors and rheumatoid arthritis. 74,75][78] This result suggests that S100a6 suppression may induce the dental epithelium to differentiate into keratinocytes rather than ameloblasts.Moreover, in the current study, S100a6 suppression strongly suppressed the matrix, but only mildly suppressed matrix-degrading enzymes (Figure 5A).It is possible that S100 family members are activated in S100a4-knockdown cells, which promote enzyme degradation.The S100 family members may regulate the expression of each other in a complementary manner.However, this study only investigated the effect of S100a6 knockdown; S100a6 suppression may be associated with the upregulation of S100a4, a8 and a9.Furthermore, the function may depend on the cell types, and there are many unknowns.A limitation of this study may be complex relationships among S100 family members.
(A) Incisors of 6-week-old WT mice.Scale bar = 500 μm.(B-D) High magnification images of each area are indicated by a white square in the low magnification images of incisors of WT mice.Scale bar = 100 μm.(E) Incisors of 6-week-old Sox21 KO mice.Scale bar = 500 μm.(F-H) High magnification images of the area enclosed by the low magnification image.Scale bar = 500 μm.Arrowheads, remaining undegraded enamel matrix; AM, apical membrane side; BM, basal membrane side.

6
Localization of calcium in ameloblasts.